Support Vector Machines for shade identification in urban areas

Abstract:- The aim of this paper is the identification of shades in urban areas for remote sensing images. The final goal consists in the risk minimization for image change detection algorithms. Correct shade identification help us to discard urban shades as urban changes. The main contribution of the paper is to focus the problem as a classification problem using the well founded Support Vector Machines theory. A comparative analysis is carried out against other classical existing classification methods where the performance of the proposed approach is verified. Key-Words:- Support vector machines, classification, density estimation